Ryan Avent on innovation

Ryan Avent has a very, very good article on the state of innovation. You should go read it. I come down on the pessimistic-but-waiting-to-be-surprised side — it really looks to me like fossil fuels were a one-time (and possibly impermanent) equilibrium-shifter — but Ryan makes a very strong case, and there’s really nothing I can disagree with.

I do have a handful of reactions, but they’re really just notes:

  • Moore’s Law is doing okay, but I’ve read a lot of people worrying that its translation into practical processing power isn’t. In general, silicon photolithography seems to have given people a lot of bad ideas about technological progression being inevitably governed by exponential laws. I’d really like to see a kW-per-dollar versus time plot of electrical engine performance, for instance (I haven’t been able to find one).

  • 3D printing is an older technology than most people realize, which makes me think they’re overestimating its upside. The medical applications are real and important; its implications for design processes are substantial; it will almost certainly become a useful neighborhood amenity, available at copy shops and the like. For manufacturing, its effects will be limited to expensive, small-run applications (like the aforementioned medical uses). It seems extremely unlikely to me that 3D printing’s economic significance will be greater than that of the 2D printer industry. Which isn’t nothing! But it’s not flying cars either.

  • Predictions of sudden algorithmic progress that unlock new types of economic activity are more interesting, and clouded by two things. First, the Big Data trend story, which is a conceptual distraction that a lot of tenuously-relevant activity is being shoehorned into. Second, the research black boxes that are the dot-com survivors. Google, in particular, has bought up a ton of the world’s best engineers and provided a sudden infusion of resources to scores of academics whose work would otherwise stand little chance of practical relevance. I think we still have only the vaguest idea of what this effort will produce. The performance of GOOG over the next few decades could turn out to be a pretty good test of Cowen’s thesis.

  • The extensive versus intensive distinction is new to me, so I’m probably about to apply it incorrectly. But it seems to me that fossil fuel technology is profoundly extensive in nature — it’s as if you can suddenly convert horses to a nonperishable liquid, and better still, there’s tons of it just lying around underground, free for the taking! These are new resources. Information technology, by contrast, is about intensifying existing activities by reducing communication friction. This is very important, but also the kind of thing that the market started working on in a serious way right after the telegraph was invented (and had been plugging away at before, of course, through couriers, postal systems, soldiers running back from the Battle of Marathon, etc). There’s an incredibly long tail to build here — smartphone-dispatched taxis, simplified appointments with the doctor, grocery delivery services you don’t have to think about — but it’s mostly composed of known tasks that aren’t sufficiently lucrative to merit hiring an unskilled human to coordinate. With enough of them, you can definitely make money. I’m not sure how economically significant it’ll wind up being, though (not least of all because consumers seem likely to capture a huge portion of the gains).

So, yeah, same old opinions from me: a lack of imagination paired with pessimism about the upside of the technologies I understand best (borne of being close enough to them that I can find their limitations dismaying).

I do think there will probably be some big unexpected thing that will make fogeys of us all. But other than driverless cars or a sudden, later-than-expected payoff to genomic medicine, I don’t see how any of the bets people are placing have enough upside to be the one that hits big.

2 Responses to “Ryan Avent on innovation”

  1. Brett

    I do think there will probably be some big unexpected thing that will make fogeys of us all.

    Artificial Intelligence?

  2. Brett

    Sorry to add-

    Although that’s not magic, like some advocates pretend. An ultra-intelligent AI can’t magic its way around physical limitations.

Leave a Reply